Artificial Intelligence... ...A Window to Mankind

Artificial Intelligence...
...A Window to Mankind
Robert Chin (Qin)
Mrs. Baron
World Lit. Honors
2/27/99

---

Qin − i
Table of Contents
Introduction......................................................................................................................1
The Difference between People and Computers.............................................................3
Logic............................................................................................................................3
Evolution.....................................................................................................................4
Turing Test..................................................................................................................4
Computers and People....................................................................................................6
Human and Computer Interaction...............................................................................6
Computers Masquerading as People..........................................................................6
Wearable Computers..................................................................................................7
Beyond Humans..............................................................................................................9
Neural Networks..........................................................................................................9
Robotics....................................................................................................................10
Expert Systems.........................................................................................................11
Agents.......................................................................................................................11
Philosophies and Moral Dilemmas.................................................................................13
Three Asmovian Laws of Robotics............................................................................13
Is there room for AI?..................................................................................................14
Free will?...................................................................................................................15
Theological outcomes of a new species of smart computers....................................16
Conclusion.....................................................................................................................18
The future uses of AI.................................................................................................18
Human society and acceptance of AI driven computers...........................................18
AI computers −− slaves or masters?.........................................................................19
Can humans eventually live forever through computers?.........................................20
Last reflections on AI and what it means to society..................................................20

---

Qin − 1
Introduction
Can machines think?
As humans, we have often asked ourselves why we are so special, why God chose to
put us here. We comb the skies looking for other signs of intelligent life through the
SETI (Search for Extra Terrestrial Intelligence) program, and try to search for the
intelligence in whales, apes and chimpanzees.
Finding consolation in none of the
above, we try to create our own intelligence, to prove to ourselves that we are not truly
alone in this universe. Perhaps it is part our search for identity that we feel we must
unlock the secrets to life, to discover whether another intelligence other than our own
can truly exist.
Science fiction.
We read it, we listen to it, and we watch it in our movie theaters.
Artificial intelligence is prevalent in most new works of science fiction, from the droid
"C3PO" in Star Wars to the cybernetic being "Data" in Star Trek.
Can this type of
artificial intelligence ever be achieved? What goes into creating such an intricate and
intelligent being?
It seems almost unfathomable that a mere machine can mimic a
human −− after all, we see computers as machines, mere tools that we utilize in our
daily routines. It is true that we are a long journey away from a truly smart computer,
but even today we make use of artificial intelligence in an ever increasing number of
applications. Robotic arms on factory assembly lines can handle more intricate tasks,
the computers in our cars respond to the road conditions and style of our driving. The
true test of human ability, however, is whether or not we can pass a computer program
off as a real live human, to fool a person into believing they are talking not to an

---

Qin − 2
electronic device, but a real human being.
To better understand artificial intelligence, I had to ask myself several questions: what
makes a person seem real? When we speak to a person, how can we tell, without
looking, that it is really a person we are talking to?
How can we simulate this?
I
contemplated these questions and realized that people go through experiences, from
which they draw on their entire life. Perhaps a truly artificial being must evolve and
learn. On the other hand, perhaps we can give an artificial being all of the memories
and thoughts that it needs. It seems highly implausible that we can reduce human
thoughts down to a set of computer algorithms that all can duplicate with ease. The
following essay will not only give the reader a better grasp of artificial intelligence, but
also will attempt to deal with the plausibility and implications of a truly artificial being.

---

Qin − 3
The Difference between People and
Computers
People and computers are vastly different in the way they think about things −− it is
fairly obvious the way a computer can perform complex mathematical calculations in the
blink of an eye while we sit pondering for years on how to factor ten digit numbers.
However computers are not perfect either.
People can look at a word problem,
decipher it, and figure out the best way to solve it. If one were to enter the exact same
information into a computer, the computer would have no idea of what to do. The ideal
problem solver would be a computer that can take advantage of both the speed of
computation of a classical computer and solve problems with the same ease as a
human. This brings us to the topic of logic.
Logic
Part of the initial problem is that computer logic is discrete, there is always a single
answer. This allows us to easily plot its decisions on a decision tree. Each node on the
tree would represent a discrete set of decisions taken, and a computer can search and
understand every single one of these decisions without actually taking them. (Johnson
39) People, on the other hand, usually make less solid decisions. For example, if we
asked whether the Rhine River, which is 820 meters in length, was long or not, we
could answer either yes or no. However, the answer is based more on opinion −− we
don’t have a fixed number in our mind that says "If the river is under 1000 meters it is
short, if it is over it is long." Depending on who we ask, we may receive a different
answer. This logic is called fuzzy logic −− it is one of the primary distinctions in the logic

---

Qin − 4
of humans and computers. (Mcneill and Freiberger 32)
Since logic stems from our
previous experiences, it only seems natural to talk about evolutionary steps next.
Evolution
It seems almost ironic that we can have the ability to create a computer program which
mimics everything a person can do −− especially since it has taken the human races a
millennia to evolve into what we are today. Through millions of years of selection, the
world has crafted humans. The ability to create a computer program that has every one
of our abilities seems almost unfathomable compared to what the human race has had
to go through. If we are to create an artificial intelligence driven computer race from
scratch, providing only the raw materials but none of the knowledge, the computer race
we create may have to undergo the same evolutionary processes in order to become
as great as ourselves. (Humphrys) Since it is impossible to measure the amount of
evolution that one creature has achieved, we must measure a computer’s intelligence
by a method which compares the computers knowledge with our own −− this test of
electronic intelligence was devised over 50 years ago −− called the Turing Test.
Turing Test
Turing was one of the greatest mathematicians. He was interested in trying to figure
out how one could measure the success of a computer program in terms of whether a
computer program has intelligence. He came up with the following rules (Hodges):
The new form of the problem can be described in terms of a game which we call the
"imitation game." It is played with three people, a man (A), a woman (B), and an

---

Qin − 5
interrogator (C) who may be of either sex. The interrogator stays in a room apart from
the other two. The object of the game for the interrogator is to determine which of the
other two is the man and which is the woman. He knows them by labels X and Y, and at
the end of the game he says either "X is A and Y is B" or "X is B and Y is A." The
interrogator is allowed to put questions to A and B.
He believed that the preceding rules would provide for the fact that if a computer’s
responses were real enough, one could not tell the difference between the computer
and the real person. Some argue, however that the turing test is simply a test of how
smart a human is, not the actual smartness of a computer program. Never the less, the
turing test remains an important key point in the development of certain AI programs −−
programs that are written with the explicit thought of trying to trick a human into
believing they are conversing with a real person while in actuality the computer program
understands none or very little of what a person is saying.
One should note that the turing test has been created, with a $100 million grand prize.
Each year, approximately 6 programs are entered in this contest −− all of which so far
have tried to win the annual $2000 prize by cheating (Hutchens and Alder) However
the turing test’s requirements are so great, that it is unlikely that the contest will push
the current envelope in AI because of the fact that many of the contest requirements for
the grand price are in technologies that haven’t been developed yet −− such as video
and audio input. Because of these requirements in advanced technologies, the most
advanced AI advances will probably come out elsewhere.

---

Qin − 6
Computers and People
Human and Computer Interaction
Humans and computers are destined to interact on a more personal level someday in
the future. As computers become smarter, they will no doubt require a more personal
interaction to achieve greater results.
No longer will we use a computer through
interface devices such as a mouse and keyboard to enter data, but instead we will ask
it to do required tasks for us, and it will do them with ease, in the same fashion that we
ourselves would. In fact, it will become feasible that computers will act and emulate
humans −− even to the fact that they will assist us in our daily tasks. In short, these
computers will end up masquerading as people.
Computers Masquerading as People
If computers are to be like us, there is no doubt that one day computers will take care of
our daily tasks transparently, calling people, arranging meetings, interacting with people
in a way that is totally undetectable. As long as one doesn’t stray too far from the
computer’s goal, the masquerading as a person is relatively easy.
This is
demonstrated through the examination of the Eliza program, a rogarian psychiatrist
designed to chat with other people. The vast number of people who chat with Eliza
believe they are being understood. The original creator of the Eliza program, Joseph
Wiezerbaum, believes that the sense a person receives that they are conversing is
created largely by the person themselves. (Weizenbaum 188) There is no doubt that
one day in the future, we may have real androids that traverse the planet, perhaps
even becoming our friends.

---

Qin − 7
It is logical that in the future there will be programs that will assist us in our daily lives
−− and completing many tasks for us. This brings us to our next topic −− augmented
reality and wearable computers.
Wearable Computers
In the future, it may be beneficial to have wearable computers. These computers would
augment reality and augment our memory. These augmenting computers would be
built with artificial intelligence in order to help us remember. Imagine having a personal
assistant who would jot notes of every interaction −− all one would need to do to recall
a previous conversation or to look up an address would be to ask the assistant. In fact,
if the assistant were intelligent enough, it would provide that information before you
even asked for it, anticipating your needs.
The most important developments in wearable computing are perhaps happening at the
MIT media lab.
There, software applications are already under development for
augmented reality and other uses. For example, imagine yourself walking down the
street. As you walk down the street, your computer projects an image, floating about
15 feet in front of you. "Today’s weather: sunny and warm with winds from the south
east at 15 mph.
Humidity 10%, Temperature 68°F, Visibility 10 miles." ("MIT:
Augmented Reality") About five minutes later, they system notifies you of new e−mail,
and prints it out onto the screen. You can read the e−mail while walking down the
street. You walk into your office, and someone comes to meet you. You’ve forgotten
their name −− but that’s okay. You’ve met them before. You touch a key, his face is
quickly scanned, recognized, and his name, address and other information is neatly

---

Qin − 8
overlaid below his head.
As you can well see, wearable computers can be very useful as supplements to our
daily lives. Currently under development is a sign language interpreter −− a computer
program that can interpret sign language for non−speakers. This could revolutionize
the way deaf people interact with society, and help to make them more accepted.
("Wearable Computer Based American Sign Language Recognizer")
Wearable computers could also be used in the workplace.
Their unique way of
overlaying text and graphics in front of a user could revolutionize the way people do
skilled labor −− the overlay could point out interesting or abnormal sections of a
product, something that may have gone unnoticed, or is totally undetectable by a
normal person. The whole point of a wearable computer is to increase the productivity
and usefulness of a human, however humans are limited in what they can do −− to
increase productivity and usefulness further, we will have to go beyond human limits.

---